CoolFace
Datasetpublic

cturan/turkish-synthetic-corpus

Turkish Synthetic Corpus A synthetic Turkish text corpus with 1,871,131 documents, designed for Turkish language model training. About Inspired by HuggingFaceTB/smollm-corpus. Questions and prompts were sourced from the SmolLM Corpus pipeline; a language model then generated localized Turkish responses and documents around them. All credit for the original corpus design and methodology goes to the HuggingFace SmolLM team. The resulting dataset covers a wide range… See the full description on the dataset page: https://huggingface.co/datasets/cturan/turkish-synthetic-corpus.

sourceHugging Faceodc-byupdated 6mo agoView on Hugging Face
0likes69downloads
Dataset Card

Turkish Synthetic Corpus

A synthetic Turkish text corpus with 1,871,131 documents, designed for Turkish language model training.

About

Inspired by HuggingFaceTB/smollm-corpus. Questions and prompts were sourced from the SmolLM Corpus pipeline; a language model then generated localized Turkish responses and documents around them. All credit for the original corpus design and methodology goes to the HuggingFace SmolLM team.

The resulting dataset covers a wide range of topics — science, history, culture, economics, coding, fiction — written in natural Turkish prose at varying register levels.

Data Fields

FieldTypeDescription
idstringUnique document ID
textstringTurkish document text

Intended Use

  • —Turkish LLM pretraining and continued pretraining
  • —Tokenizer training
  • —Language modeling benchmarks
Content is synthetic. Factual accuracy is not guaranteed.

Citation

If you use this dataset, please also credit the original SmolLM Corpus:

bibtex
@software{benallal2024smollmcorpus,
  author = {Ben Allal, Loubna and Lozhkov, Anton and Penedo, Guilherme and Wolf, Thomas and von Werra, Leandro},
  title = {SmolLM-Corpus},
  month = July,
  year = 2024,
  url = {https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus}
}